快讯
谷歌FlowAgent:在代码评审流程中自动修复测试失败
谷歌论文提出FlowAgent,在CI中以ReAct式"生成-验证"循环处理预提交测试失败,并将修复展示在代码评审工具中;两道弃权过滤器分别在执行前后拦截弱建议。在195个真实失败的人工评审中,67.18%的修复正确;全公司上线后共为295,508个变更建议修复,开发者预览65,069个、采纳28,554个。
事件来源
查看原文
RT elvis Great paper from Google on program repair agents inside CI. (bookmark it) They propose FlowAgent, which runs a ReAct-style generate-and-validate loop on pre-submit test failures and shows the fix inside Google's code review tools. Two abstention filters decide whether a fix is worth showing. One runs before execution and one after, so weak suggestions never reach the developer. In a manual review of 195 real failures, 67.18% of its fixes were correct. After the Google-wide launch, it suggested fixes on 295,508 changes. Developers previewed 65,069 and applied 28,554. Paper: https://arxiv.org/abs/2610.07289 Chat with Paper: https://academy.dair.ai/papers/catching-developers-in-the-flow-low-latency-agentic-program-repair-at-google-sca-2610.07289